CN105095065A - Optimization method for formalized modeling - Google Patents

Abstract

The present invention relates to an optimization method for formalized modeling. In a critical and real-time flight control system that has strict requirements on time, single use of a formalized modeling design method causes time cost brought by partial intermediate redundant codes generated in a modeling process and requirements of a hard real-time task system cannot be met. According to the scheme adopted by the present invention, by introducing a formalized modeling tool and a bridging technological method of a general library, an architecture design of a system is completed by means of a formalized modeling method at an early period and calculation analysis is performed on performances, such as time cost, of each module in the system; and for the module that has relatively large time cost or not good performance, an existing general library is bridged by using a self-adaptive adaptation layer, so that cooperative work of the formalized modeling method and the general library is implemented, and performance optimization is performed on formalized modeling.

Description

A kind of optimization method of Formal Modeling

Technical field

The invention belongs to flight control system software design arts.

Background technology

At flight control system Software for Design development field, main designing and developing comprises h coding's development approach and Formal Modeling method for designing.

Traditional flight control system software development is mainly by h coding's software design, can the general-purpose library of generating portion maturation by the exploitation of multiple model and iteration, these general-purpose libraries, through the checking of multiple model, have the compact and efficient feature of code.If but system requirements change, h coding is not easy to the maintenance that demand changes, and also lacks the support of patterned emulation and verification tool in the later stage simultaneously.

Along with the development of flight control system gordian technique, start to introduce Formal Modeling in increasing aerial key modules, software requirement analysis is realized by different forms modeling, Formal Modeling is based on very strict mathematical theory, by the strict modeling to requirement profile, remove ambiguity and the ambiguity of system requirements.Portion-form modeling tool can automatic code generating, and code generator have passed civil aviaton's standard, makes to develop software while very much not spending and shortening the development time, ensures the security of system well.

But contain a large amount of redundancy intermediate codes in the code due to generation automatically in Formal Modeling, and flight control system real-time task is very harsh to the requirement of time margin, General Requirements time margin needs 30 percent of the time being greater than the whole cycle, and expense extra time that these redundancy intermediate codes are brought, cause formal Verification Techniques to appoint the application in periodic duty very limited at the hard real time of key.

Summary of the invention

The object of the invention is the designing technique introduced based on Formal Modeling and general-purpose library bridge joint, this technology, by building self-adaptation adaptation layer, solves the problem that general-purpose library effectively can not be applied in Formal Modeling.

The technical scheme that the present invention takes is

An optimization method for Formal Modeling, is characterized in that,

Step 1 builds the architectural framework of system by Formal Modeling instrument;

Step 2 carries out performance analysis and test for the module in framework, filters out the object element needing improving performance;

Step 3 is by building self-adaptation adaptation layer, and the unit higher with performance in general-purpose library to object element mates and replace;

Architectural framework after optimization carries out verifying and emulates by step 4 under Formal Modeling platform.

Described self-adaptation adaptation layer is built as follows,

The data-interface of object element and the data-interface of general-purpose library in the modeling of first step contrast version, complete data-interface coupling;

Second step completes Formal Modeling object element and mates with the logic of the higher unit of performance in general-purpose library.

The advantage that the present invention has and beneficial effect: the present invention is a kind of optimization method to Formal Modeling, by building self-adaptation adaptation layer, unit not good for performance in Formal Modeling framework is carried out screening strength, and in general-purpose library, choose that code is compact, checking fully, the high unit little with time overhead of operational efficiency replace, thus both can possess Formal Modeling automatically advantage fast, turn avoid the performance issues such as the time overhead that automatic code generating brings.Call previously ripe general-purpose library by self-adaptation adaptation layer, carried out significantly optimizing to the performance of Formal Modeling, also overcome Formal Modeling simultaneously and cannot be applied to the shortcoming in the crucial real-time system of the flight of time requirement harshness,

Accompanying drawing explanation

Fig. 1 is the optimization method schematic flow sheet of fault recovery Formal Modeling

Fig. 2 is the Formal Modeling architectural framework of Failure Recovery Module

Fig. 3 is fault recovery general-purpose library block schematic illustration

Fig. 4 is fault recovery self-adaptation adaptation layer schematic diagram

Embodiment

Embodiment

Below for the Failure Recovery Module in flight control system software, by reference to the accompanying drawings, be introduced a kind of Formal Modeling optimization method, overview flow chart is shown in Fig. 1, and concrete steps are as follows:

System requirement analysis is carried out in step a pair fault recovery, and in the fault recovery of flight control system, the deciding grade and level that can recover fault and the fault the recovered project defined, should limit cautiously and to some extent.When primary fault appears in recoverable signal, answer lighting to point out pilot to carry out failure recovery operation, then determine whether recover by pilot, and implement the control of recovery by pilot;

Step 2 carries out Architecture Analysis to fault recovery, carries out functional analysis and Formal Modeling analysis, and utilizes the KCG in SCADE instrument to complete real time workshop, as shown in Figure 2 by the state machine (SSM) under SCADE environment to fault recovery;

Step 3 analysis of failure recovers the complexity of each module under framework, and test failure recovers the time overhead of each module under architectural framework, find that the operation (FailRecvFuncFSM) of Petri Nets state belongs to the larger unit of time overhead, air tasking time requirement can not be met;

Step 4 is by carrying out analyzing to existing fault recovery software generic storehouse and screening, refer to Fig. 3, by analyzing complexity and time overhead, filter out following three functions alternatively change the larger Petri Nets state of time overhead under modeling under operation part, and performance index and the requirement of system can be met by the following function unit in the mode determination fault recovery general-purpose library of computing time:

Fault recovery sensor process function (Lib_AnaFailRcvRetHandler)

Fault recovery discrete magnitude process function (Lib_DisFailRcvRetHandler)

Fault recovery result treatment function (Lib_FailRcvRetProc)

The function declaration operated under the step 5 KCG instrument analyzed under SCADE environment generates Petri Nets state automatically, by the situation of the parameter type and number and rreturn value of analyzing this function, the power function of the Petri Nets operation in more common storehouse again, component self-adaptation adaptation layer:

1. newly-built FailRecvAdaptor.c file, realizes the self-adaptation adaption function of data-interface by definition self-adaptation fitness function (FailRecvParamAdaptor);

voidAdaptor_FailRecvParam(inC_FSM*inC,outC_FSM*outC，FailRecvFsmIn*v_st_fsm_in)

2. the time overhead operating (FailRecvFuncFSM) analyzed Petri Nets state in step 3 under is larger, hereof according to the statement of the fault recovery state processing function under Formal Modeling, defined function: voidFailRecvFuncFSM (inC_FSM*inC, outC_FSM*outC);

3. in step 4, filtered out three functions as replacement function, function prototype is as follows: fault recovery sensor process function

DT_UINT16Lib_AnaFailRcvRetHandler (FailRecvFsmIn*v_st_fsm_in) fault recovery discrete magnitude process function

DT_UINT16Lib_DisFailRcvRetHandler (FailRecvFsmIn*v_st_fsm_in) fault recovery result treatment function

DT_UINT16Lib_FailRcvRetProc(FailRecvFsmIn*v_st_fsm_in)

4. under completing Formal Modeling Petri Nets state, operation is adaptive with the function of general-purpose library Petri Nets, refers to Fig. 4;

The object element of the integrated Formal Modeling architectural framework of step 6 and general-purpose library, carries out emulating based on the Simulation instrument under SACDE environment and verifies.

Claims (2)

1. an optimization method for Formal Modeling, is characterized in that,

Step 1 builds the architectural framework of system by Formal Modeling instrument;

Step 2 carries out performance analysis and test for the module in framework, filters out the object element needing improving performance;

Step 3 is by building self-adaptation adaptation layer, and the unit higher with performance in general-purpose library to object element mates and replace;

Architectural framework after optimization carries out verifying and emulates by step 4 under Formal Modeling platform.

2. the optimization method of Formal Modeling as claimed in claim 1, it is characterized in that, described self-adaptation adaptation layer is built as follows,

The data-interface of object element and the data-interface of general-purpose library in the modeling of first step contrast version, complete data-interface coupling;

Second step completes Formal Modeling object element and mates with the logic of the higher unit of performance in general-purpose library.